Britain - The Official Thread

[Robin]

That needs to drop a bit further. Me and the missus want to go to Scotland later this year.

They will be using a different currency by then!

 

[Touring Mars]

A few interesting points from the first two results of the evening:

Turnout is looking higher than 2015 - up to 10% higher.
UKIP are taking a hammering, but UKIP votes are going to both Labour and Conservative.
Sunderland's election results appeared to be read out by a Dalek.

 

[Famine]

Turnout is looking higher than 2015 - up to 10% higher.

I found the calls to get out and vote because high turnout hurts the Tories interesting.

Since the days of Thatcher, Labour has only won one election with more than 62% turnout. And that was 1997.

 

[Liquid]

My prediction was no great change in seats with Labour only increasing their vote in seats they already hold.

That was before the exit poll. Now, my own former constituency is predicted to go Tory; it's been a Labour seat since 1945. This would be a huge change alongside the 5-6 other predicted Tory seats in North Wales.

 

[Famine]

Interesting comment from the very first declared seat:

According to the exit poll there should be a 5 point decrease in the Conservative vote here, while Labour’s vote is expected to be up by 19 points.

In practice, the Conservative vote is up by 6 points and the Labour vote by 10.

The result is thus a smaller swing to Labour than expected by the exit poll.

 

[Liquid]

I've said for a while that the Tories will still surprise even if they don't win seats.

Silent voters are usually Tory, in my opinion.

 

[Famine]

And Sunderland South:

According to the exit poll there should be a 4 point increase in the Conservative vote here, while Labour’s vote is expected to be up by 13 points.

In practice, the Conservative vote is up by 11 points and the Labour vote by just 4.

The result is thus a much better result for the Conservatives than expected by the exit poll.

The exit polls have never been wrong by more than 20 seats for any party...

 

[Liquid]

All the North East Wales seats just got prime coverage on the BBC's show.

Alyn & Deeside
Delyn
Wrexham
Clwyd South

All going Tory. This is a simply enormous change for the region. And not one I would enjoy.

Edit: Third seat in. Tories getting good votes even in a seat they don't win again. Doing better than EP suggests. Again.

 

[Famine]

All the North East Wales seats just got prime coverage on the BBC's show.

Alyn & Deeside
Delyn
Wrexham
Clwyd South

All going Tory. This is a simply enormous change for the region. And not one I would enjoy.

Either I don't understand the exit polls or they're bigly wrong.

How can a party that's pencilled in to lose 10% of its support and 5% of its seats be making these gains?

 

[Liquid]

Either I don't understand the exit polls or they're bigly wrong.

How can a party that's pencilled in to lose 10% of its support and 5% of its seats be making these gains?

People lying at exit polls?

Exit polls also do not account for postal votes. Just something to bear in mind.

 

[Sanji Himura]

Either I don't understand the exit polls or they're bigly wrong.

How can a party that's pencilled in to lose 10% of its support and 5% of its seats be making these gains?

We could front the same question as to why the Republicans retained Congress even though they lost seats in the last election.

 

[PeterJB]

People lying at exit polls?

Exit polls also do not account for postal votes. Just something to bear in mind.

Are there enough postal votes to make that much of a difference?

 

[Famine]

People lying at exit polls?

Or small samples (30k though - still that's only 50 per constituency) and from the wrong places?

Sunderland Central is Labour hold, but 2.3% swing to the Tories.

I almost daren't go to bed at this point...

 

[TRGTspecialist]

If there was going to be a year where even the exit poll couldn't figure what's going on, it would be this one...........

I noted that when John Curtice was speaking earlier he was implying that how constituencies voted in the EU referendum was factoring heavily into how they judged the exit poll. Seems a reasonable assumption but not every election/by-election that's happened since has followed the pattern of "Leave = voting Tory/UKIP, Remain = voting Labour/Lib Dem".

If in the minds of voters this wasn't a "Brexit election" then that could lead to bigger errors than normal.

For what it's worth a lot of the Twitterati still remain convinced that it is wrong.

 

[Liquid]

Or small samples (30k though - still that's only 50 per constituency) and from the wrong places?

30,450 sample for an electorate of 46.9 million is what... less than 0.1%? It's far too small to go off.

But if it is true, and the Tories do sweep North East Wales, they have to be losing big time elsewhere. Somewhere.

Apparently Home Secretary Amber Rudd's seat is under threat so there could be at least one Portillo moment tonight.

 

[TRGTspecialist]

30,450 sample for an electorate of 46.9 million is what... less than 0.1%? It's far too small to go off.

I don't recall the maths but sample sizes of thousands are more powerful than you think (opinion polls generally work with just 1,000-2,000, and even Britain's poor polling gets within 4-6% of the result on average).

Plus that sample won't be from every constituency; they'll be taken from a relatively small number of marginal seats, and the results of all the safer seats will be projected from those findings.

 

[Milouse]

I don't recall the maths but sample sizes of thousands are more powerful than you think

maybe you have this in mind : https://www.gtplanet.net/forum/threads/europe-the-official-thread.230488/page-30#post-11808034

Anyway, for this kind of election we had today in UK (and sunday in France), it's very hard to extrapolates as we're talking about 650 elections, not one.

Edit: question: why don't we have almost full results yet, it's 1h09 AM (i guess 23pm or midnight in UK), did poll booth close very late?

 

[TenEightyOne]

Either I don't understand the exit polls or they're bigly wrong.

They only represent a sample of those who voted at the polling station(s) for that ward, they exclude the postal vote. Areas with (comparatively) high student or BAF populations are often not-so-well represented in exit polls.

30,450 sample for an electorate of 46.9 million is what... less than 0.1%? It's far too small to go off.

It's often far more accurate than the lengthy pre-vote polls, the mechanism is quite clever for the most part, particularly the "swing" calculations. See my comment to Famine above for caveats.

I note as the evening goes on that there are surprising swings both ways in England and Wales, I think the exit poll is slightly unfavourable to my predicted eventual actuality for the Conservatives but it's not going to be the resounding mandate that Theresa May was seeking.

The Conservatives under Ruth Davies appear to be making mighty gains in Scotland at the SNP's expense. Many of the changes across England, Scotland and Wales are along the Leave/Remain lines reflected in the Brexit Referendum. That's clearly the big issue for many voters.

Liberals contributing to a coalition? The party leadership and committee remain resolute that that won't happen. I think they may as well do it if the opportunity arises - things can hardly get worse for their voter standing and they have the opportunity to directly affect Brexit policy, something which is very dear to them.

BREAKING: Amber Rudd may lose her seat. That's something at least

 

[TRGTspecialist]

Edit: question: why don't we have almost full results yet, it's 1h09 AM (i guess 23pm or midnight in UK), did poll booth close very late?

Yeah, polls closed at 10pm (2 and a half hours ago).

 

[Milouse]

It's a 1 Round election it seems, what mechanism keep the voters "voice" to be fragmented in too much candidate? Can you have an unlimited numbers of candidates in an area (one of 650 district), and/or different parties represented in various area?
I have 14 candidates to choose from for sunday's parlimentary election in France, but there's a 2nd round with remaining ones. But in UK i don't know how it works with only one round.

 

[jake2013guy]

It's a 1 Round election it seems, what mechanism keep the voters "voice" to be fragmented in too much candidate? Can you have an unlimited numbers of candidates in an area (one of 650 district), and/or different parties represented in various area?
I have 14 candidates to choose from for sunday's parlimentary election in France, but there's a 2nd round with remaining ones. But in UK i don't know how it works with only one round.

There are a number of parties who choose to stand in a constituency. The party with the most votes gets that constituency (seat) and the party with the most seats is the leading party. And whoever has a majority (Over 326 in this instance) holds the most representatives in Parliament.

 

[TRGTspecialist]

Very early days but here is how the exit poll (profiled only to seats that have been called) stacks up against actual seat results so far.


https://twitter.com/maitlis/status/872960093519835136
But worth noting that the Tories' overperformance so far has come from safe Labour seats, where it doesn't matter - in more marginal seats they've been doing worse.

It's a 1 Round election it seems, what mechanism keep the voters "voice" to be fragmented in too much candidate? Can you have an unlimited numbers of candidates in an area (one of 650 district), and/or different parties represented in various area?
I have 14 candidates to choose from for sunday's parlimentary election in France, but there's a 2nd round with remaining ones. But in UK i don't know how it works with only one round.

Yep, our system is first-past-the-post, so whoever gets a plurality of votes in each constituency is elected, regardless of whether they have a majority or not.

There's no limit on how many candidates can run or how much the vote can be split - I believe the record is held by South Belfast in the 2015 election, where less than a quarter of voters chose the elected MP!

 

[TenEightyOne]

It's a 1 Round election it seems, what mechanism keep the voters "voice" to be fragmented in too much candidate? Can you have an unlimited numbers of candidates in an area (one of 650 district), and/or different parties represented in various area?
I have 14 candidates to choose from for sunday's parlimentary election in France, but there's a 2nd round with remaining ones. But in UK i don't know how it works with only one round.

Each area (or 'constituency') has a list of candidates, this can be any number of candidates although it's normally around 5 or 6. The candidate with the most votes in that constituency takes a seat in the House of Commons. The party that takes the most seats is "the winner"... but they have to have more seats than all their opponents put together to form a majority.

Exit polls look at each constituency and measure the swing of support from/to the most popular parties (in most cases that's Conservatives or Labour) and predict the result from that. In that way you don't need to know the voting turnout to predict how the vote will fall. It's still not an exact science, naturally, but in recent elections it's given far more accurate results than the pre-election polls. For parties in a constituency who get a small percentage of the overall area's vote the swing projection is far less accurate for obvious reasons.

Majority news: DUP will ally with the Conservatives if necessary, no real surprises there.

 

[Liquid]

Labour hold Wrexham by ~2,000 votes. That is a blow for the Welsh Tories no matter what else happens; Wrexham is the biggest and most powerful place in North Wales.

Welsh Tories tried desperately to secure Wrexham. Is it a coincidence that Theresa May actually went there yet still couldn't do it?

 

[TRGTspecialist]

First gain of the night - Labour gain Rutherglen & Hamilton West from the SNP, 9% swing. Suggests that the mini-collapse for the SNP predicted by the exit poll is real.

 

[jake2013guy]

Corbyn is 10/11 favourite.

 

[TRGTspecialist]

Ok, Tories are doing bad enough in the marginals (and very badly in London) it looks possible that the exit poll has overestimated their seat share.

Sinn Fein could be taking up to 7 seats, and since they don't go to Westminster they reduce the number of MPs required to form a majority. The speaker (who only votes for the status quo if there is a tie) can be discounted as well. So that makes the working majority 322 MPs.

The DUP, who I think are the only party who will be willing to support the Tories, may take up to 10 seats. So I reckon that the Tories need to take no less than 312 MPs to have a chance of forming a government.

EDIT: The one place where the Tories are doing unexpectedly well and gaining seats is, bizarrely, Scotland. BBC have upgraded their exit poll to a forecast, which now puts the Tories on 322, Labour on 261.

 

[Liquid]

Some face on him.

 

[ECGadget]

Well... Sheffield have just had a Cleggxit.

 

[Crumpet]